WORKFORCE PROJECTIONS AND JOB MARKET TRENDS FOR GRADUATE AND PROFESSIONAL DEGREE RECIPIENTS

Transcription

1 EDUCATING THE NEXT GENERATION OF CALIFORNIANS IN A RESEARCH UNIVERSITY CONTEXT: University of California Graduate and Undergraduate Enrollment Planning Through 2010 Appendix 2 WORKFORCE PROJECTIONS AND JOB MARKET TRENDS FOR GRADUATE AND PROFESSIONAL DEGREE RECIPIENTS Planning and Analysis Academic Affairs Office of the President University of California February 1999

2 TABLE OF CONTENTS Page EXECUTIVE SUMMARY..... iii INTRODUCTION I. DIFFICULTIES IN PROJECTING FUTURE WORKFORCE NEEDS... 2 II. PH.D. WORKFORCE NEEDS... 5 A. Broad Job Market Trends and Projections for Ph.D. Recipients... 5 Overview of Ph.D. Job Market Trends in the 1990s... 5 Multi-Field Projections of Future Ph.D. Workforce Needs... 8 Bowen and Sosa (1989) Atkinson (1990) COSEPUP (1995) Massy and Goldman (1995) Dauffenbach (1996) U.S. Bureau of Labor Statistics (1997 and 1998) Changing Conditions B. Ph.D. Employment Trends and Outlooks, By Discipline Engineering and Computer Sciences Computer sciences, engineering (overall), chemical engineering, civil engineering, electrical engineering, mechanical engineering Life Sciences Biomedical/other biological sciences Physical Sciences and Mathematics Chemistry, mathematics, physics Social Sciences and Psychology Anthropology, economics, history, political science, psychology, sociology Humanities and Arts English, foreign languages and literature, philosophy, fine and applied arts Table 1: Recent Employment Trends and Projections for Ph.D. Scientists and Engineers, by Discipline, Selected Sources.. 34

3 III. WORKFORCE NEEDS FOR PROFESSIONAL AND MASTERS DEGREE RECIPIENTS A. Engineering and Computer Scientists (Masters-Level) Computer scientists and systems analysts, engineers B. Other Professionals Business executives, K-12 teachers, lawyers, other professionals, new professionals and other masters degree recipients IV. BEST GUESSES: WORKFORCE NEEDS FOR UC GRADUATE AND PROFESSIONAL DEGREE RECIPIENTS ii

4 EXECUTIVE SUMMARY Under the California Master Plan for Higher Education, the University of California has a responsibility to prepare graduate and professional degree recipients to help meet California s and the nation s workforce needs. Workforce needs, therefore, are an important criterion in planning UC s graduate and professional enrollments. As part of the analysis for the University s long-range enrollment planning through the year 2010, currently in progress, this background report analyzes recent labor market trends and projected workforce needs for doctoral, professional, and masters degree recipients where possible, by occupation and disciplinary field. Projecting future workforce needs for graduate and professional degree recipients is a challenging task for several reasons. We cannot assume that future workforce needs in six or more years (when students now beginning doctoral study will enter the labor market) will mirror today s needs. Moreover, projecting needs even a few years in the future for professionals in fields closely tied to the economy, such as business and law, is difficult. Current job market conditions, as well as our own projections about future needs, influence the decisions that students, educators, and policy makers make about graduate study, thereby altering the supply of individuals entering the workforce. The labor market outlook also varies substantially by field. Finally, graduate and professional degree recipients not only fill existing workforce needs but shape the workforce, creating new demand and new opportunities for economic growth and social benefits through the jobs they take. Indeed, a review of major national Ph.D. workforce studies and projections developed over the past decade indicates that different analyses have reached sometimes conflicting conclusions regarding future Ph.D. demand and supply, with some analyses projecting large Ph.D. shortages and others projecting large Ph.D. surpluses. The most recent projections present a somewhat more optimistic outlook than do those developed during the economic recession of the early 1990s. Despite the difficulty in projecting workforce needs, we believe they are too important not to consider in UC s enrollment planning. We therefore have tried to summarize and evaluate available information and to provide our best guesses on needs for UC graduate and professional degree recipients, but we recognize that our ability to project the future accurately is limited. Consequently, workforce analysis will require continuous environmental monitoring and student and employer feedback, and planning assumptions will have to be revisited regularly, as new information becomes available. In providing our best guesses about future workforce needs, it is important to note two points. First, California s economic, social, demographic, and environmental needs and outlook will differ from the nation as a whole. Although UC prepares graduate students who will meet not just state but national and international labor market needs, many UC graduates (especially those from professional programs) remain in California. Over the next decade, according to economic analysts, the California economy, especially that related to high-tech industries, is expected to be stronger than that for the U.S. as a whole, and college enrollments will grow at a faster rate. California is expected to outpace the nation in jobs, income, and population growth. iii

5 Second, demand for graduates of UC s graduate and professional programs may be greater than that for advanced degree holders nationally, because of the quality and focus of UC s programs. UC graduate and professional degree recipients generally have had good placement records, even when the job market has been relatively weak, and UC Ph.D. s overall have better placement records than do Ph.D. s nationally, especially in engineering/computer sciences and physical sciences/mathematics. Given the information available at this time regarding job market trends, workforce projections, and continuing uncertainties, as well as the distinctive needs of California and the success of UC graduates, our best guesses about future workforce needs over the next decade for graduate and professional degree recipients from the University of California are as follows: Ph.D. recipients: We expect that UC will need to increase the number of doctoral students it prepares, to fill three main types of positions: (1) college and university faculty, to meet demand due to enrollment growth and faculty replacement needs, (2) private-sector scientists and engineers, especially in high-tech areas, and (3) nontraditional and alternative Ph.D. careers, to fill and create fields and occupations that are just now emerging, for example, in new interdisciplinary fields and new applications to existing fields. Based on information currently available, we expect that needs will be strongest in fields such as computer sciences, engineering, public administration, and some applied psychology areas. Privatesector demand for Ph.D. s in chemistry and the biological sciences, as well as in computer sciences and engineering, is expected to increase significantly, especially in growing hightech industries; this demand likely will be even stronger in California. Conversely, supply may continue to outpace demand in mathematics and in a number of humanities and social science fields. Nevertheless, within each of these disciplines, there will likely be demand in particular areas of specialization. Professional and masters degree recipients: Professional enrollments are closely tied to California s needs, since most UC-educated professionals remain in the state. To meet California s expanding and changing economic and social service needs, the numbers of UC professional and masters degree enrollments will need to increase, especially in computer science and engineering fields linked to high-tech industries. Demand for K-12 teachers with the kind of strong academic training that UC can provide will also be high. While demand for business executives and lawyers is expected to grow, it is unclear whether supply will outpace demand. Finally, there may be significant needs for new types of professionals with an interdisciplinary masters-level training that builds on UC s research base, especially in emerging technology-oriented fields (e.g., digital arts) or in areas of growing social concern, such as environmental studies. iv

6 INTRODUCTION 1 Under the California Master Plan for Higher Education, the University of California has a responsibility to prepare graduate and professional degree recipients capable of meeting not just current but changing workforce needs for California and the nation. Workforce needs, therefore, are an important criterion in planning UC s graduate and professional enrollments. As part of the analysis for the University s long-range enrollment planning through the year 2010, currently in progress, this background report analyzes recent labor market trends and projected workforce needs for doctoral, professional, and masters degree recipients in California and the U.S. where possible, by occupation and disciplinary field. We hope that the report, especially information by specific occupations and disciplinary fields, will also be useful to campuses and programs in their enrollment planning. However, determining the numbers of individuals the workforce will need as well as the numbers that UC and other institutions will enroll over the next decade is a challenging task. Consequently, workforce analysis will require continuous environmental monitoring and student and employer feedback, and enrollment planning will require institutional flexibility in thinking about how best to meet these needs. The report is organized as follows: Section I presents a brief overview of the challenges inherent in projecting future workforce needs. Section II focuses on the labor market outlook for Ph.D. recipients. 2 This section first reviews recent data and studies that can inform our understanding of broad job market trends, broad workforce projections for doctoral recipients, and changing workforce conditions. Since job trends and outlooks for doctoral recipients vary widely by discipline, we then examine job market trends and projections for a number of specific disciplines for which information was available. Section III analyzes job market trends and projected workforce needs for professional and masters degree recipients in a number of fields. Finally, Section IV summarizes our best guesses about future workforce needs for graduate and professional degree recipients, focusing specifically on the role that we expect UC programs to play in meeting these needs. 1 This analysis builds upon earlier analyses developed by the Office of Planning and Analysis in the University of California Office of the President, as well as upon a chapter by Ami Zusman, Issues Facing Higher Education in the Twenty-First Century, in Philip G. Altbach, Robert O. Berdahl, and Patricia J. Gumport, eds., American Higher Education in the Twenty-first Century: Social, Political, and Economic Challenges (Baltimore: The Johns Hopkins University Press, 1999). 2 The term Ph.D. in this report also includes other academic doctorates (e.g., Ed.D., D. Engin.).

7 I. DIFFICULTIES IN PROJECTING FUTURE WORKFORCE NEEDS Determining future workforce needs for graduate and professional degree recipients is a challenging task for several reasons. We cannot assume that future workforce needs will mirror today s needs. For example, the employment outlook for Ph.D.'s in six or more years (when students now beginning doctoral study will enter the job market) may be far different than it is now. For college faculty, we may need to project at least eight years into the future. Most life science and physical science Ph.D. recipients complete their degrees in about six years but then complete two or more years of postdoctoral work before they are hired as faculty. In the social sciences and humanities, Ph.D. students typically take seven to eight years to complete their degrees but then generally move directly into faculty positions. Given recent economic volatility, projecting needs even a few years in the future for professionals in fields closely tied to the economy, such as business and law, is difficult. In addition, as always, there will be shifts in demand for professionals that cannot be fully anticipated. The narrower the focus (i.e., specific disciplines or subfields), the greater the chance that projections based on current employment conditions will err. Fluctuations in demand for individual disciplines are even greater than that for the overall Ph.D. market or for broad fields. Yet to plan graduate enrollments and programs, we need information by discipline. Current job market conditions, as well as the projections we make about future workforce needs, influence individuals decisions, thereby altering the future demand/supply ratio. As Richard Freeman noted as far back as the 1970s, a boom and bust cycle exists in the academic labor market for Ph.D. s: When current Ph.D. jobs appear plentiful, growing numbers of individuals apply to graduate programs, but fewer apply when jobs are in short supply. In addition, our own projections of future demand influence student decisions about whether or not to enter doctoral programs and institutional and governmental decisions about whether or not to expand and support them; in so doing, these decisions may invalidate the projections. Because a lag exists between job market needs and Ph.D. production, job shortages are inevitably followed by surpluses and surpluses by shortages. 3 Demand for graduates of UC s graduate and professional degree programs may differ from national prospects. California s economic, social, demographic, and environmental needs and outlook will differ from the nation as a whole. Although UC prepares graduate students who will meet not just state but national and international labor market needs, many UC graduates remain in California. This is especially true of graduates of UC s professional programs, the great majority of whom work in California; but many Ph.D. recipients, especially in engineering 3 Richard Freeman, The Overeducated American (New York: Academic Press, 1976). 2

8 and the physical sciences, take jobs in California as well. 4 Over the next decade, according to economic analysts, the California economy, especially that related to high-tech industries, is expected to be stronger than that for the U.S. as a whole, and college enrollments will grow at a faster rate. California is expected to outpace the nation in jobs, income, and population growth. 5 Demand for graduate and professional degree recipients from UC may be greater than that for advanced degree holders nationally because of the quality and focus of UC s programs. UC graduate and professional degree recipients generally have had good placement records, even during periods of a relatively weak job market. (See Section IV.) Graduate and professional degree recipients themselves shape workforce needs and requirements. Graduate and professional degree recipients not only fill existing workforce needs but shape the workforce, creating new demand and new opportunities for economic growth and social benefits through the jobs they take. Masters and Ph.D. recipients in nontraditional careers that might once have been considered "out-of-field" or inappropriate have transformed those positions by bringing their skills and knowledge to bear on these positions, so that now holding the advanced degree is a job requirement. For example, those entering public-sector policy analysis are now often required to hold an advanced degree, although this was not true years ago. Experience with the skillful analysis advanced degree holders provide has increased the profession s expectations for the kind of analyses that will be carried out and for the kinds of skills that incumbents will bring to their tasks. Given these and other challenges in projecting future graduate and professional demand and supply, some analysts have concluded that it is not now possible, and may never be possible, to forecast future workforce needs for doctoral and other advanced degree recipients with accuracy sufficient for planning, and that we therefore should not attempt to do so. 6 Rather, such observers would base graduate-level enrollments on criteria such as student and program quality, avoid sharp enrollment increases or decreases, and modify graduate programs to provide greater breadth and flexibility, better career guidance, and financial support mechanisms that encourage 4 For example, among recent UC engineering doctorates, over 72 percent sought immediate employment (rather than postdoctoral appointments) after graduating; of these, nearly 60 percent obtained jobs in California upon graduating. Nearly 50 percent of physics Ph.D. s seeking immediate employment and over 40 percent of chemistry Ph.D. s seeking employment obtained jobs in California (National Research Council, Survey of Earned Doctorates, 1995, UC data files). Many Ph.D. s who accept postdoctoral appointments ultimately will take career positions in California colleges, industry, or other sectors. 5 Center for Continuing Study of the California Economy (CCSCE), California Economic Growth, 1998 Edition (Palo Alto, CA: CCSCE, 1998). 6 Ad Hoc Working Group on the Supply of Science, Engineering, and Mathematics (SEM) Professionals, Federal Coordinating Council for Science, Engineering, and Technology, Report (September 9, 1993); National Science Foundation, Division of Science Resources Studies, Who Is Unemployed? Factors Affecting Unemployment Among Individuals with Doctoral Degrees in Science and Engineering, Special Report, NSF , By Carolyn F. Shettle (Arlington, VA: NSF, 1997). 3

9 student flexibility in meeting employment demands. 7 High quality must, of course, be a prerequisite for all UC graduate programs and enrollments, and flexibility and responsiveness are essential. However, despite the difficulty in projecting workforce needs, we believe they are too important not to consider in UC s enrollment planning. We therefore have tried to evaluate the best information available, but we recognize that planning assumptions will have to be revisited regularly, as new information becomes available. 7 Committee on Science, Engineering, and Public Policy (COSEPUP) of the National Academy of Sciences, the National Academy of Engineering, and the Institute of Medicine, Reshaping the Graduate Education of Scientists and Engineers (Washington, D.C.: National Academy Press, 1995). Henceforth, we refer to this report as the COSEPUP report. 4

10 II. PH.D. WORKFORCE NEEDS This section first reviews recent data and studies that can inform our understanding of broad job market trends, broad workforce projections for doctoral recipients, and changing workforce conditions. Since job trends and outlooks for doctoral recipients vary widely by discipline, we then examine job market trends and projections for a number of specific disciplines for which information was available. A. Broad Job Market Trends and Projections for Ph.D. Recipients Overview of Ph.D. Job Market Trends in the 1990s In the early 1990s, the Ph.D. job market dropped to a 20-year low point, a fact that generated widespread concerns about Ph.D. employment and the size of graduate programs. In the first half of the 1990s, the job market for most U.S. Ph.D. recipients was worse than it had been for at least two decades. This was true not only for Ph.D. s in the humanities and social sciences, many of whom have long faced lengthy job searches after graduating, but for new engineering and physical science Ph.D. s as well. The poor job market for Ph.D. s (as well as for many professionals) was due to a number of unanticipated changes, including perhaps most importantly the economic recession of the early 1990s. As a result both of the recession and of shifts in state tax dollars toward growing demand for services in areas such as health, corrections, and K-12 education, funding for public higher education in California and in many other states was constrained and cut, sometimes severely, and public colleges hired fewer new faculty than they had earlier expected. In addition, the end of the Cold War brought about downsizing in the defense industry, which employs many doctoral scientists and engineers. Other reductions in private-sector research and development (R&D) in business and industry and in government employment also hurt job prospects for new Ph.D. s. At the same time, partly in response to earlier predictions of Ph.D. job shortages, doctoral production (which had been expected to remain level) increased more than 30 percent between 1986 and Nationally, nearly half of all physical scientists and over 70 percent of biological science doctorates now seek postdoctoral scholar appointments after graduating, rather than immediate 8 U.S. Department of Education, National Center for Education Statistics (NCES), Projections of Education Statistics to 2008, NCES , by Debra E. Gerald and William J. Hussar (Washington, D.C.: U.S. Government Printing Office, 1998). 5

11 employment, 9 and the number of years they spend in postdoctoral positions appears to be lengthening. These increases reflect the emergence of a period of mentored postdoctoral research and study as an expected part of the preparation of scientists, but the extended number of years that new Ph.D. s are spending in such appointments is also a response in part to a weaker job market in the sciences. 10 Part-time, out-of-field, or temporary employment among new Ph.D. s rose in the early 1990s, as well. In the humanities, 14 percent of Ph.D. s were involuntarily working part-time or outside their fields in April 1993, and 11 percent were in temporary positions. 11 Even during weak job markets, however, unemployment among Ph.D. s has been considerably below that of the labor force as a whole. Even during the difficult years in the early 1990s, most Ph.D. s, in both science and non-science fields, eventually found jobs, and unemployment for Ph.D. s remained considerably lower than that for the U.S. civilian labor force as a whole. For instance, in 1993, the overall U.S. unemployment rate was 6.8 percent. By contrast, the overall unemployment rate in 1993 for recent science, engineering, and social science Ph.D. s (that is, Ph.D. s who had received their degrees one to three years earlier) was 1.7 percent, and it was 3.3 percent for recent humanities Ph.D. s. In 1995, overall unemployment for recent science, engineering, and social science Ph.D. s was 1.9 percent, ranging from 0.5 percent in psychology to 4.3 percent in chemical engineering, and it was 3.0 percent in the humanities. 12 In recent years the job market for Ph.D. s in most fields has begun to improve, especially for jobs in the industrial sector. Some industries have created new opportunities for Ph.D. holders in recent years. Most U.S. science and engineering Ph.D. s and nearly half of all social science doctorates no longer work in four-year colleges or universities, the traditional employer of Ph.D. s. Rather, by 1991, business and industry had become the largest single employment sector for both engineering and physical science Ph.D. s. As the recent COSEPUP report concluded, PhDs are increasingly finding employment outside universities and more and more are in types of positions that they had not expected to occupy National Research Council, Summary Report: Doctoral Recipients from United States Universities, 1996, by Peter H. Henderson, Julie E. Clarke, and Cynthia Woods (Washington, D.C.: National Academy Press, 1998). Physical science figure includes mathematics but excludes computer sciences (which UC classifies with engineering/computer sciences). Even higher proportions of UC Ph.D. s in the physical sciences (as well as in engineering) seek postdoctoral scholar appointments as a step toward career positions 57 percent of physical science/mathematics Ph.D. s, and 27 percent of Ph.D. s in engineering/computer sciences. 10 Commission on Professionals in Science and Technology (CPST), Postdocs and Career Prospects: A Status Report (Washington, D.C.: American Association for the Advancement of Science, 1997). 11 National Science Board, Science & Engineering Indicators 1996 (Washington, D.C.: U.S. Government Printing Office, 1996). 12 Ibid.; Mark Regets, What s Happening in the Labor Market for Recent Science and Engineering Ph.D. Recipients?, National Science Foundation Issue Brief, NSF (Washington, D.C.: NSF, September 23, 1997); National Research Council, Humanities Doctorates in the United States: 1995 Profile, by Linda Ingram and Prudence Brown (Washington, D.C., National Academy Press, 1997). 13 COSEPUP, Reshaping the Graduate Education of Scientists and Engineers. 6

12 Moreover, as noted further below, recent surveys by a number of professional associations, including chemistry, history, languages and literature, mathematics, and physics, conclude that the Ph.D. job market for new Ph.D. s (i.e., those graduating within the previous year) is improving, especially in the industrial sector. In addition, a survey of a sample of new doctorates in 13 science, social science, and engineering fields found that unemployment as of mid-october 1997 ranged from 0.6 percent in psychology to 4.6 percent in chemistry. 14 Since unemployment levels are usually highest in the first few months after Ph.D. s graduate, when they are first seeking positions, we would expect that unemployment rates for these individuals have dropped still further since October Nevertheless, job market difficulties for Ph.D. s in some fields remain. Despite evidence of an improved job market, several cautions should be kept in mind. First, these are improvements over what in many cases had been the worst job market in two decades. In a number of fields, including mathematics, English, and history, the supply of new Ph.D. s continues to outpace demand, and many new Ph.D. s in these fields are still having difficulty finding jobs in their fields, especially potentially permanent positions. For example, a 1997 report by the Modern Language Association concluded that new English and foreign language Ph.D. s continue to face a crisis in the academic job market (where most such Ph.D. s are employed) because the number of Ph.D. s awarded annually in these fields continues to rise faster than full-time faculty openings. 15 Second, in some cases, faculty hires are coming from a backlog of Ph.D. recipients who had been in temporary positions, while new Ph.D. s continue to enter the job market. For example, although the number of newly hired assistant professors in history rose substantially between 1996 and 1997, over a third of them had received their degrees more than three years earlier. 16 Third, as noted, larger numbers of Ph.D. s in the sciences are spending longer periods of time in temporary postdoctoral appointments. According to surveys by a number of professional associations, many postdocs, especially those in fields like chemistry, earth sciences, mathematics, and physics, say they are taking such temporary positions because they were unable to find permanent jobs. 17 Finally, in fields such as political science, mathematics, 14 CPST, Employment of Recent Doctoral Graduates in S&E: Results of Professional Society Surveys (Washington, D.C.: CPST, 1998). Percentages in this report should be viewed with some caution, because of low sample size in some fields, low response rates (ranging from 34 percent in engineering to 73 percent in physiology), exclusion of those with foreign addresses in physics, and other problems that may bias the responses. (The CPST report includes a fourteenth field, political science, with seven percent unemployment; however, CPST advises interpreting this rate cautiously, because of the particularly small sample size.) 15 Modern Language Association of America, MLA Committee on Professional Employment, Final Report (Sandra Gilbert, Chair) (MLA: December 1997). 16 Robert B. Townsend, Academic Job Opportunities Better than Expected in 1997, American Historical Association, Perspectives (October 1997): 9, CPST, Postdocs and Career Prospects; CPST, Employment of Recent Doctoral Graduates in S&E. By contrast, taking a postdoctoral position has become an expected career step in some fields, especially in the biological sciences. 7

13 sociology/anthropology, and the humanities, many employed Ph.D. s continue to be involuntarily employed part-time or outside their fields. 18 Multi-Field Projections of Future Ph.D. Workforce Needs National Ph.D. workforce projections paint a mixed picture, with differences not only by field but among different labor force analysts. The most recent projections present a somewhat more optimistic outlook than do those developed during the economic recession of the early 1990s. In spite of the difficulties associated with producing projections, several labor force economists and other scholars, governmental agencies, and professional associations have attempted in recent years to assess the nation s future workforce needs for doctoral recipients. A number of these analyses are broad-based studies that cut across a range of fields. These studies provide important sources of information regarding overall Ph.D. labor market needs, and several have had a significant influence on public discussions about future Ph.D. needs. Therefore, it is important both to note their conclusions and to understand their methods and assumptions. There is no clear consensus among these studies about future workforce needs for doctoral recipients, however. Indeed, these analyses disagree not only about the scope and nature of such future workforce needs but whether there will be shortages or surpluses. Rather, their conclusions appear to be influenced by the assumptions and methodologies used, as well as by the conditions existing at the time in which they were developed. For example, those developed in the late 1980s, prior to the economic recession and the end of the Cold War, projected substantial shortages of Ph.D. s in many fields, while those developed in the midst of the recession made more pessimistic assumptions. Each study has significant limitations and, in hindsight, did not anticipate important recent developments. Yet, they remain our best source of information about future Ph.D. workforce needs as well as underscoring the difficulties inherent in attempting to project the future and, thus, the need to ensure that we continue to monitor employment conditions and to maintain flexibility to adapt programs and enrollments to changing needs. We summarize briefly below, in chronological order, the major studies of projected Ph.D. workforce needs that have been conducted over the past decade. 1. William G. Bowen and Julie Ann Sosa, Prospects for Faculty in the Arts and Sciences: A Study of Factors Affecting Demand and Supply, 1987 to 2012 (Princeton, N.J.: Princeton University Press, 1989). (Focus: arts and sciences faculty for four-year colleges.) In what remains the most recent comprehensive set of labor market projections for college faculty in arts and sciences fields, Bowen and Sosa projected serious faculty shortages after 1997 across the arts and sciences, with particularly severe shortages in the humanities and social sciences during the period. Basing their analysis on conditions and trends present in 18 Mark Regets, What s Happening in the Labor Market for Recent Science and Engineering Ph.D. Recipients? ; National Science Board, Science and Engineering Indicators Data on involuntary out-of-field or part-time employment date from 1995 or earlier; the stronger economy in the past three years may have reduced these numbers. 8

14 the mid-1980s, Bowen and Sosa projected substantial faculty shortages as a result of several expected outcomes: large numbers of faculty would retire and need to be replaced after 1997, as faculty hired in the 1960s and 1970s reached retirement age; college enrollments would increase; and the number of new doctorates seeking academic careers would decline, falling well short of demand between 1997 and They also assumed that the arts and sciences student-faculty ratio at four-year colleges in the mid-1980s (then at what may have been an all-time low) would be maintained or even decline further and that the time the typical graduate student took to complete the doctorate would continue to increase. Although the authors presented four alternative models of faculty demand, the highest and lowest of these projections differed by only eight percent. To date, these projections of serious faculty shortages have not come to pass and seem unlikely at least in the next few years. New doctorates soared 30 percent between 1986 and 1996, rather than declining. In addition, because colleges and universities suffered sharp budgetary cutbacks in the 1990s, they hired fewer new faculty than anticipated. In the humanities and in some of the physical sciences, a reserve pool of recent Ph.D.'s in temporary positions are eager to assume faculty positions. Moreover, the authors' assumption that the low student-faculty ratio would be maintained or decline even further seems unrealistic for the foreseeable future. A crude calculation of full-time student-faculty ratios (for all fields) in four-year institutions nationally suggests that at best they have not changed in recent years; in California, they have risen. Moreover, even though the total number of college faculty grew nearly 13 percent between 1991 and 1995 (with higher growth rates in two-year colleges), the number of full-time faculty increased less than three percent, and the number of newly hired full-time faculty declined nearly eight percent. 19 However, because faculty retirements are likely to grow in the coming years, there may be greater numbers of new hires in the future. Nevertheless, from the vantage point of 1999, Bowen and Sosa s projections of future Ph.D. supply to 2007 seem too low by far, while their projections of future faculty demand seem overly optimistic, even at a time of anticipated enrollment increases, faculty retirements, and improved economic conditions. 2. Richard C. Atkinson, Supply and Demand for Scientists and Engineers: A National Crisis in the Making, Science, 248 (27 April 1990): pp (Focus: natural science and engineering Ph.D. s.) Based on data from a draft National Science Foundation study, Atkinson, then president of the American Association for the Advancement of Science and chancellor of UC San Diego, projected a critical shortfall of Ph.D. scientists and engineers for both academic and privatesector positions by the early years of the 21 st century. He calculated that an annual shortfall of several thousand doctoral scientists and engineers would develop over the next several years and would persist well into the 21 st century. Three factors in particular were expected to increase demand for Ph.D. scientists and engineers: increasing faculty retirements, which would require faculty replacements, growth in college enrollments, and anticipated growth in federal and private R&D investments, resulting in industry need for additional Ph.D. scientists and engineers. Like the Bowen and Sosa study, the NSF study projected that supply would fall short of demand. The NSF study projected that the number of new Ph.D.'s would peak in 1993 and 19 U.S. Department of Education, NCES, E.D. TABS: Fall Staff in Postsecondary Institutions, 1995, NCES , by Stephen Roey and Rebecca Rak (Washington, D.C.: U.S. Government Printing Office, 1998). 9

15 then decline nearly 17 percent by 2003, before subsequently beginning to recover. In the early 1990s, the unpublished NSF study was widely cited as evidence for the need to intervene to increase the number of individuals pursuing Ph.D.'s in the sciences and engineering. The NSF study appears to have underestimated the numbers of students already in the doctoral pipeline in In addition, calls to increase the number of Ph.D.'s further may have been successful. By 1993, the number of new science and engineering doctorates awarded was 26 percent higher than projected and, rather than declining after 1993, the number of new science and engineering Ph.D.'s increased another eight percent between 1993 and This increase, together with the unanticipated economic recession and the defense downsizing, has meant that the projected shortfall has not materialized to date. On the other hand, perhaps in response to widespread reports about a weaker doctoral job market, applications to graduate programs nationally began to decline in a number of fields beginning in 1993, leading to enrollment declines. 20 In particular, first-time graduate enrollments in the physical sciences and engineering declined by 15 and 20 percent, respectively, between 1991 and 1996; only in 1997 did these numbers level off or increase slightly. 21 If new graduate enrollments remain at these lower levels or continue to decline, the number of new Ph.D.'s in these fields may decline in the next few years. Nevertheless, in retrospect this study, too, appears to have overestimated projected shortages at least for the next few years, although current graduate enrollment patterns and an improving R&D environment could produce Ph.D. shortages further down the road Committee on Science, Engineering, and Public Policy (COSEPUP) of the National Academy of Sciences, the National Academy of Engineering, and the Institute of Medicine, Reshaping the Graduate Education of Scientists and Engineers (Washington, D.C.: National Academy Press, 1995). (Focus: Ph.D. s in the physical sciences, life sciences, social sciences, and engineering.) The COSEPUP report found that, while new Ph.D. s in the sciences, social sciences, and engineering were facing increasing difficulties in obtaining positions during the weak doctoral job market of the early 1990s, most eventually did obtain jobs. The report concluded that the overall demand for Ph.D. scientists and engineers remains strong but that the types of positions for which these Ph.D.'s will be needed are changing. In general, the COSEPUP report expected slowed growth in academic positions and reduced demand for traditional basic researchers in some fields (both within and outside academe) but a relatively good market for those seeking applied R&D positions in business and industry, as well as non-research positions, especially for those who are adaptable and can bring diverse skills: "Over the long term, demand for graduate 20 Peter D. Syverson, Early Returns of 1997 CGS/GRE Survey Reveal Second Year of Graduate Enrollment Decreases, CGS Communicator (Newsletter of the Council of Graduate Schools), XXXI (November/December 1998): pp National Science Foundation, Graduate Students and Postdoctorates in Science and Engineering: Fall 1997 Supplemental Tables. Data are for full-time students in both masters and doctoral programs; it is not known what proportion of the decline is at the doctoral (rather than masters) level. It should also be noted that first-time graduate enrollments in many fields, as well as first-time undergraduate enrollments, declined in the early 1990s. 22 The UC situation with regard to applications to and enrollments in graduate engineering and computer science programs differs somewhat from the national one. While applications to such UC programs declined from 1993 through 1995, they rose four percent in 1996 and another six percent between 1996 and These UC programs have now increased their admissions in response to needs, and new enrollments have increased as well, returning the enrollment level for new students to that of the early 1990s. 10

16 scientists and engineers in business and industry is increasing; more employment options are available to graduate scientists and engineers who have multiple disciplines, minor degrees, personal communication skills, and entrepreneurial initiative" (p. 2-17). The report did not make specific workforce projections, arguing that current supply-demand models could not adequately predict undersupply or oversupply of scientists and engineers, in part because students, faculty, policy makers and others alter their behavior on the basis of employment projections, thereby altering labor supply outcomes. However, the report found no convincing arguments to limit the number of graduate enrollments. Rather, it suggested that universities broaden graduate training to prepare individuals for a wider range of prospective careers. Critics have argued that the report is overly optimistic because it discounts individuals who may be employed but may not be in permanent career positions appropriate to their training. 4. William F. Massy and Charles A. Goldman, The Production and Utilization of Science and Engineering Doctorates in the United States (Stanford, CA: Stanford Institute for Higher Education Research Discussion Paper, May 1995, with August 1995 revisions). (Focus: Ph.D. s in the physical sciences, life sciences, social sciences, and engineering.) Using econometric model simulations of supply and demand, Massy and Goldman projected employment gaps in 12 fields physics/astronomy, chemistry, bioscience, geological sciences, mathematics, economics, psychology, computer science, and four engineering fields. Massy and Goldman concluded that, given conditions prevailing in the early 1990s, about 22 percent of science and engineering doctorates in the U.S. could fail to find suitable employment, with wide variations among fields. For example, they projected labor surpluses ranging from a 44 percent surplus in mechanical engineering to no or negligible surpluses in chemistry, psychology, and computer science. Many higher education experts have argued that the study's assumptions and methodology, and therefore its conclusions, are flawed. 23 For example, the model ignored differences among fields in industrial demand and in retirement rates; engineering fields were projected to have high labor surpluses, in part because the model failed to consider the higher rates of industry demand for engineering Ph.D. s. The model also is based on several questionable assumptions, for example, assumptions of a steady-state in employment demand and supply, steady-state undergraduate enrollments, and assumptions about the proportions of foreign students who will remain in the U.S. Nor did the model account for major changes in the National Research Council s definitions of field of study and employment, which resulted in some cases in a narrow definition of appropriate field of employment; for example, medical science positions are excluded from the workforce demand for bioscience Ph.D. s. Among the model's surprising conclusions are an estimated labor shortage in chemistry (five percent) and large labor surpluses in electrical engineering (41 percent) and bioscience (28 percent) contrary to conclusions drawn by most professional association surveys and other studies. Nevertheless, the report received wide attention from both policy makers and the media. Our evaluation of the study is that, while it is an impressive effort to model complex variables, the measures used in the simulations do not 23 See, for example, Peter D. Syverson, When Simulation Becomes Reality: Press Reaction to Massy/Goldman Study Creates Erroneous Message, CGS Communicator (Newsletter of the Council of Graduate Schools) (August 1995): 8-10; Charlotte Kuh, Is There a Ph.D. Glut? Is that the Right Question?, CGS Communicator (August/September 1996): 1,

17 reflect reality, as a comparison between the model s projections and actual situations shows. Consequently, we have not found the study helpful in projecting future Ph.D. demand and supply. 5. Robert C. Dauffenbach, mid-growth scenario for supply and demand of scientists and engineers (1994 to 2005), unpublished tabulations, cited by the National Science Board, Science and Engineering Indicators 1996 (Washington, D.C.: U.S. Government Printing Office, 1996), Chapter 3. (Focus: scientists and engineers, all degree levels.) Under a mid-growth scenario in a model developed by Robert C. Dauffenbach, supply and demand for scientists and engineers (not separated by degree level) will remain more or less in balance between 1994 and Labor shortages are projected in a few areas (especially for computer and mathematics specialists and, to a smaller extent, engineers), and slight labor surpluses are projected in other areas, such as for physical scientists. Much of the growth in demand is not projected to occur until after It is important to emphasize that Dauffenbach s model does not distinguish between demand or supply at different degree levels (baccalaureate, masters, and doctoral) and that only about five percent of the science and engineering workforce has doctoral degrees. The National Science Board notes that it does not endorse this or other models but presents it as one possible approach worthy of examination. 6. U.S. Bureau of Labor Statistics, Occupational Outlook Handbook (Washington, D.C.: U.S. Government Printing Office, 1998); George T. Silvestri, Occupational Employment Projections to 2006, U.S. Bureau of Labor Statistics, Monthly Labor Review, 120 (November 1997): pp ; George T. Silvestri, Occupational Employment to 2005, U.S. Bureau of Labor Statistics, Monthly Labor Review, 118 (November 1995): pp (Focus: two- and four-year college faculty, all fields; also doctoral degree recipients, all fields.) College and university faculty: The most recent U.S. Bureau of Labor Statistics (BLS) projections for college faculty demand and supply, which provide projections for the period from 1996 through 2006, are more optimistic than previous reports. The Occupational Outlook Handbook projects that there will be many openings for college faculty (at two- or four-year institutions) during the 10-year projection period, as large numbers of faculty members retire, beginning in the late 1990s. In addition to replacements, the total numbers of college faculty will increase about as fast as the average for all occupations (i.e., a 10 to 20 percent increase), as college enrollments increase. Faculty prospects will continue to be better in certain fields (such as business, engineering, health science, and computer science) where applicants also have attractive job opportunities outside academe. On an annual basis, the BLS s November 1997 projections of total job openings for college faculty through 2006 (including both growth in jobs and replacements) are nearly 15 percent higher than November 1995 projections through (In fact, however, the BLS notes that job openings are expected to be greater in the later years than in the earlier years.) However, the Handbook predicts keen competition for full-time college faculty positions, as growing numbers of new Ph.D. s compete for what is expected to be a decreasing number of such full-time positions. In addition, even though demand may increase, in some fields the supply of Ph.D. s will increase faster than demand. In terms of projecting needs for doctoral recipients, one shortcoming of the BLS projections is that they combine job 12

18 openings for faculty in four-year institutions (where most faculty have doctorates) with those in two-year institutions (where only a small proportion of faculty currently hold doctorates). Doctoral degree recipients: According to BLS projections, total job openings between 1996 and 2006, as a proportion of 1996 employment, will be greater for doctoral degree recipients (45 percent) than for all occupations (38 percent). (Job openings for bachelors recipients will be 46 percent of 1996 employment, while those for first-professional (law and health professions) and masters degree recipients will be 34 percent and 31 percent, respectively.) ******** In summary, these various national projections of Ph.D. workforce demand and supply provide us some, but limited, guidance in assessing the workforce component of UC doctoral enrollment planning. Projections of both large Ph.D. shortages (for example, the Bowen and Sosa and Atkinson studies) and large surpluses (for example, the Massy and Goldman study) seem less likely to be realized than do projections of moderate growth in demand. The most recent projections suggest that much of the growth in such demand, where it occurs, is not likely for several years. Since entry of Ph.D. s into the workforce generally takes six years or more from the start of doctoral study, this time frame fits with our enrollment planning, but it means that many Ph.D. s currently in the pipeline may continue to face a somewhat difficult job market. As detailed in Section IIB below, it is important to emphasize that Ph.D. demand and supply will vary greatly by discipline. But year-to-year fluctuations in demand for Ph.D. s in particular disciplines are even greater than those in the overall Ph.D. market or in broad fields, and unanticipated events (for example, economic recession or substantially increased federal investment in research and development) could radically alter demand and supply. This again points to the need for constant monitoring and flexibility. 24 Changing Conditions If economic conditions remain strong, overall demand for Ph.D. s will likely increase. Less certain is whether Ph.D. supply (the numbers of new Ph.D. s awarded nationally) will match, fall short, or exceed demand. While the workforce projections published to date help guide our thinking, we must pay special attention, in planning future UC graduate enrollments, to several important and changing conditions that will affect Ph.D. workforce needs. Some of these changing conditions will likely increase workforce needs for doctoral recipients; others will likely constrain them. These factors will impact individual disciplines differently. 24 Additional studies now underway should provide further information in the next year or two on the Ph.D. labor market and our ability to project it. For example, the National Academy of Sciences has established a Committee on Methods of Forecasting Demand and Supply of Doctoral Scientists and Engineers, to examine additional forecasting methodologies and recommend improvements in the way that forecasts are presented. A national study on Ph.D. s Ten Years Later, being conducted by Maresi Nerad and Joseph Cerny of UC Berkeley, will provide better understanding of long-term Ph.D. career patterns in six representative disciplines, and a special substudy of Ph.D. s from the nine UC campuses will provide specific information on how UC Ph.D. recipients have fared. 13

19 Enrollment growth: California college enrollments are expected to increase significantly over the next decade, as Tidal Wave II (the large numbers of students now in the secondary school system) enters college in the early years of the 21 st century. In the next seven years alone, UC s preliminary planning estimates project an increase of about 15 percent between and from 153,600 to about 176,000; and CSU enrollments are projected to increase over 23 percent from 350,900 to 432,000, according to 1998 Department of Finance projections. 25 Although debate continues regarding the specifics of different enrollment projections, there is no question that enrollments in California s public four-year institutions will increase significantly. 26 If, as expected, enrollments in CSU and UC grow by approximately 104,000 over the next seven years, roughly 5,200 to 5,550 new faculty could be required, with still more faculty needed as enrollments continue to grow beyond Nationally, enrollments in four-year institutions are projected to increase more slowly but still substantially during this seven-year period an eight percent increase from 8.9 million to 9.6 million students, with faster enrollment growth after This enrollment growth will require colleges and universities to hire more Ph.D. faculty, although the fields in which new faculty will be in heaviest demand are not fully certain. (We address this question in more detail below.) In addition, if more working adults return to school for advanced education, new opportunities for faculty hiring may open. Faculty retirements: Large numbers of faculty, both in California and in the U.S., were hired in the 1960s to teach the baby boom generation, which not only enrolled in college in large numbers but also enrolled in higher proportions than their predecessors. As many of these faculty reach retirement age in the next decade, colleges will be hiring new faculty to replace them. Nationally, nearly half (44 percent) of all full-time faculty in the U.S. were aged 50 or older as of At the California State University, which has nearly 11,000 full-time faculty, 58 percent of full-time faculty were 50 years or older in 1997; 16 percent were 60 years or older. At UC, 45 percent of tenure-track faculty were 50 years or older in Consequently, colleges and universities may be looking to replace close to half of their current faculties in the next 10 to 15 years. 29 The level of faculty hiring: Even though enrollments will increase and large numbers of faculty will retire over the next decade, it is possible that colleges and universities will not 25 UC figures are based on year-average headcount enrollments and preliminary projections. CSU figures are projected Fall headcount enrollments, as projected by the California Department of Finance, Demographic Research Unit in its California Public Postsecondary Enrollment Projections, 1998 Series. 26 We limited our analysis of state and national enrollment projections to four-year colleges and universities, because the great majority of faculty in these institutions has Ph.D. degrees. By contrast, only about 17 percent of full-time two-year college faculty in the U.S. (or roughly three to four percent of all full-time faculty) hold doctoral degrees. 27 This assumes student-faculty ratios no lower than 18.7:1 or higher than 20:1. 28 U.S. Department of Education, NCES, Projections of Education Statistics to Middle alternative projections for enrollments in 4-year institutions. 29 U.S. Department of Education, NCES, Digest of Education Statistics (Washington, D.C.: U.S. Government Printing Office, 1997); CSU, Faculty and staff data, Fall 1997; UC, Academic Personnel Report 9, October

20 hire new faculty in the numbers once expected. A number of factors could affect faculty hiring levels. Institutions might further raise their student-faculty ratios if there are continuing or recurring constraints on public funding, along with constraints on tuition increases. Indeed, between 1991 and 1995, many of the positions vacated by retiring faculty in a number of science and engineering fields, including the social sciences, mathematics, physical sciences, and psychology, were not filled by new faculty; and overall numbers of faculty declined. It is too soon to know whether this is a long-term shift in faculty hiring patterns or a temporary one, in response to the economic recession of the early 1990s and the decline in overall college enrollments. 30 There has also been speculation that greater use of distance-learning and other instructional technology delivery systems could reduce the number of regular faculty that institutions hire. In addition, if college access and affordability decline, college-going rates may be lower than now projected. Also, the traditional college-age population will grow more slowly after about Finally, if many older faculty choose to stay on beyond traditional retirement age, if Ph.D.'s now in temporary positions succeed in finding permanent positions, or if there is an influx of immigrant scientists and engineers, there could be fewer openings for new doctorates. Industry opportunities: The resurgent California and U.S. economies will likely require growing numbers of Ph.D. s and other advanced degree recipients, especially in engineering and the sciences, to meet industry R&D needs. The private sector, primarily industry, now employs well over half of all engineering and physical science Ph.D.'s in the U.S., and the proportion of life science Ph.D.'s employed in industry is growing, especially in areas such as biotechnology and pharmaceuticals. 31 If another recession should occur, industry R&D could go through another round of downsizing, similar to that which occurred in the late 1980s and early 1990s. On the other hand, if the economy remains strong, as most forecasters expect, industrial demand for Ph.D. scientists and engineers is expected to increase significantly, especially in areas such as electronics, biotechnology, and other hightech areas, but with more of these opportunities in development and applied research work. Ph.D. supply: Whether there will be too many or too few Ph.D.'s to meet workforce needs also depends on the number of new doctorates awarded. Universities around the country have been responding to many of the same trends affecting California, including declines or stagnation in state government support, anticipation until recently of significant cuts in federal research support for graduate research assistantships, and concerns about a weak job market. As has occurred at UC, other institutions have reduced graduate admissions in some departments in order to ensure that those students who do enroll have adequate financial support and in response to concerns about the job market. Prospective graduate students also have responded to combined concerns about the job market and the adequacy of financial support, including rising debt burdens, by choosing not to enroll. Between 1992 and 1997, first-time full-time graduate enrollments in the sciences, social sciences, and engineering 30 Rolf F. Lehming, What is Happening to Academic Employment of Scientists and Engineers?, National Science Foundation Issue Brief, NSF (Washington, D.C.: NSF, June 19, 1998). 31 Keith Wilkinson, Number of Doctoral Scientists and Engineers Grows by 6 Percent between 1993 and 1995, National Science Foundation Data Brief, NSF (Washington, D.C.: NSF, March 13, 1997, as revised October 16, 1997). 15